communication security for smart grid distribution...

IEEE Communications Magazine • January 201342 0163-6804/13/$25.00 © 2013 IEEE

INTRODUCTION

The current electrical grid is perhaps the great-est engineering achievement of the 20th century.However, it is increasingly outdated and over-burdened, leading to costly blackouts andburnouts. For this and various other reasons,transformation efforts are underway to make thecurrent electrical grid smarter.

The smart grid could be referred to as themodernization of the current electric grid for thepurpose of enabling bidirectional flows of infor-mation and electricity in order to achieve numer-ous goals; it will provide consumers with diversechoices on how, when, and how much electricity

they use. It is self-healing in case of distur-bances, such as physical and cyber attacks andnatural disasters. Moreover, the smart grid’sinfrastructure will be able to link and utilize awide array of energy sources including renew-able energy producers and mobile energy stor-age. Additionally, this infrastructure aims atproviding better power quality and more effi-cient delivery of electricity. Indeed, all thesegoals could not be achieved and realized withouta communication technology infrastructure thatwill gather, assemble, and synthesize data pro-vided by smart meters, electrical vehicles, sen-sors, and computer and information technologysystems.

CYBER SECURITY MOTIVATIONHistory has proven that industrial control sys-tems were in fact vulnerable to and victims ofcyber attacks. In March 2007, Idaho NationalLaboratory conducted an experiment in whichphysical damage was caused to a diesel genera-tor through the exploitation of a security flaw inits control system. Additionally, during the Rus-sian-Georgian war in 2008, cyber attacks widelybelieved to have originated in Russia broughtdown the Georgian electric grid during the Rus-sian army’s advance through the country. Besidesthat, in April 2009, the Wall Street Journal report-ed that cyber spies had penetrated the U.S. elec-trical grid and left behind software programsthat could be used to disrupt the system. Lastbut very significant, in 2010, Stuxnet, a largecomplex piece of malware with many differentcomponents and functionalities, targetedSiemens industrial control systems and exploitedfour zero-day vulnerabilities running Windowsoperating systems. As a result, 60 percent of Ira-nian nuclear infrastructure was targeted, hencetriggering genuine fear over the commencementof cyber warfare.

It is therefore of utmost importance toaddress the cyber security aspect of the smartgrid, specifically the area concerned with thecommunication mechanisms that deal with thedistribution subpart.

The rest of the article is organized as follows.We pinpoint some related work in our area ofconcern, and then illustrate and describe thesmart grid architecture. We thoroughly elaborate

ABSTRACT

The operation and control of the next gener-ation electrical grids will depend on a complexnetwork of computers, software, and communi-cation technologies. Being compromised by amalicious adversary would cause significant dam-age, including extended power outages anddestruction of electrical equipment. Moreover,the implementation of the smart grid will includethe deployment of many new enabling technolo-gies such as advanced sensors and metering, andthe integration of distributed generationresources. Such technologies and various otherswill require the addition and utilization of multi-ple communication mechanisms and infrastruc-tures that may suffer from serious cybervulnerabilities. These need to be addressed inorder to increase the security and thus the great-est adoption and success of the smart grid. Inthis article, we focus on the communicationsecurity aspect, which deals with the distributioncomponent of the smart grid. Consequently, wetarget the network security of the advancedmetering infrastructure coupled with the datacommunication toward the transmission infra-structure. We discuss the security and feasibilityaspects of possible communication mechanismsthat could be adopted on that subpart of thegrid. By accomplishing this, the correlated vul-nerabilities in these systems could be remediat-ed, and associated risks may be mitigated for thepurpose of enhancing the cyber security of thefuture electric grid.

CYBER SECURITY FOR SMART GRIDCOMMUNICATIONS: PART 2

Elias Bou-Harb, Claude Fachkha, Makan Pourzandi, Mourad Debbabi, and Chadi Assi, Concordia University

Communication Security for Smart GridDistribution Networks

ASSI LAYOUT_Layout 1 12/21/12 12:01 PM Page 42

IEEE Communications Magazine • January 2013 43

on the feasible communication mechanisms inthe distribution part of the smart grid, revealingtheir security objectives, security threats, andpractically applicable implementation on thefuture grid. We present a discussion of the secu-rity framework that is needed to enable thosecommunication techniques. Finally, we summa-rize and conclude this article.

RELATED WORKIn this section, we briefly highlight some of thework done in the communications and securityarea in the context of smart grid distribution.

Metke et al. [1] discussed key security tech-nologies for a smart grid system including publickey infrastructures (PKI) and trusted computingfor various smart grid communication networks.They thoroughly presented the security require-ments that are essential for the proper operationof the future grid. In another research work, Yuet al. [2] identified the fundamental challenges indata communications for the smart grid andintroduced the ongoing standardization effort inthe industry. Moreover, the authors depicted thecommunication infrastructures — home areanetworks (HANs) and neighborhood area net-works (NANs) — and very briefly listed themechanisms utilized to achieve their architec-tures. In another article entitled “Secure Com-munications in the Smart Grid” [3], the authorsfocused on HANs by elaborating on their appli-cation/manager interface (AMI) infrastructure,

and its security issues and requirements. Theauthors expressed their model in terms of asecure communication mechanism on that sub-part of the grid.

To the best of our knowledge, the work beingpresented in this article is unique in providingsignificant, relevant, and practical informationon the communication mechanisms in bothHANs and NANs, by focusing on their security,including their objectives and threats, in additionto their practical feasibility, requirements, andsecurity issues when implemented on the smartgrid.

SMART GRID ARCHITECTUREIn this section, we provide a high-level overviewof the architecture of the smart grid as depictedin Fig. 1. The future electric grid has a tieredarchitecture to supply energy to consumers.Energy starts from power generation and flowsthrough transmission systems to distribution andeventually to consumers. The smart grid is striv-ing to utilize and coordinate various generationand production mechanisms. Moreover, genera-tion plants can be mobile or fixed depending onspecific architectures. On the transmission side,a large number of substations and network oper-ating centers manage this task. A large numberof mixed voltage power lines transmit the gener-ated electricity from various sources to the dis-tribution architecture. Finally, a set of complexdistribution topologies delivers the electricity to

Figure 1. Smart grid architecture.

EV charging

Residential areas

Rural farm

Metropolitan areas

Manufacturing factory

DistributionTransmissionProduction

High voltageSolar power plant

Hydro power plant

Wind power plant

Coalpower plant

Mobile plant

Lowvoltage

Distributionsubstation

Nuclear powerplant

Transmissionsubstation

Networkoperatingcenter

Extra highvoltage

Rural farm

History has proven

that industrial control

systems were in fact

vulnerable and vic-

tims of cyber attacks.

It is therefore of

utmost importance

to address the cyber

security aspect of the

smart grid, specifical-

ly the area con-

cerned with the

communication

mechanisms which

deal with the distri-

bution subpart.


IEEE Communications Magazine • January 201344

regions, neighbors, and premises for utilizationand consumption.

In this article, our interest lies in the distribu-tion part of the smart grid. More specifically, weare concerned with the communication networksof that subpart of the grid, the HAN and theNAN. These networks are critical for data com-munications between the utility and end users.HANs are composed of three components: first,the smart in-house devices that provide demand-side management such as energy efficiency man-agement and demand response; second, thesmart meter that collects data from smart devicesand invokes certain actions depending on theinformation it retrieves from the grid; and third,the HAN gateway, which refers to the functionthat links the HAN with the NAN. This gatewaycan also represent the physical device dedicatedto performing this functionality. On the otherhand, a NAN connects multiple HANs to localaccess points where transmission lines carry thedata out toward the utility.

COMMUNICATION MECHANISMSIn this section, we focus on the communicationsecurity aspect that deals with the distributionand consumption components of the smart grid.In the remainder of this section, we follow thesubsequent methodology. First, we pinpoint themost applicable and utilized communicationmechanisms that could be adopted on that sub-part of the grid by introducing their technologyand use. Second, we discuss their security objec-tives including confidentiality, integrity, authenti-cation, and authorization. Third, we elaborateon their threats and vulnerabilities. Finally, wediscuss their feasibility in the context of theirimplementation and security on smart gridHANs and NANs.

HAN COMMUNICATION MECHANISMSThe AMI is the key element in smart grid HANs[4]. It is dubbed the convergence of the powergrid, the communication infrastructure, and thesupporting information architecture. It refers tothe systems that measure, collect, and analyzeenergy usage from advanced smart devices,including in-home devices as well as electricvehicle charging, through various communicationmedia, for the purpose of forwarding the data tothe grid. Thus, this critical communication infra-structure ought to be discussed and investigated.

Wireless LAN — 802.11 is a set of standardsdeveloped for wireless local area networks(WLANs). It specifies an interface between awireless device and a base station (access point)or between two wireless devices (peer-to-peer).

802.11 provides confidentiality by implement-ing the advanced encryption standard (AES).Integrity is achieved through the AES-CBC-MAC algorithm [5], while authentication isimplemented using the Wi-Fi Protected Accessstandards. IEEE 802.11 by default does not offerauthorization mechanisms.

The protocol suffers from significant securitythreats. It is vulnerable to traffic analysis, a tech-nique that allows the attacker to determine theload on the communication medium by monitor-

ing and analyzing the number and size of pack-ets being transmitted. It is also susceptible topassive and active eavesdropping where anattacker can listen to the wireless connection aswell as actively inject messages into the commu-nication medium. Moreover, 802.11 is vulnerableto man-in-the-middle, session hijacking, andreplay attacks.

On one hand, it can be declared that theWLAN (802.11) technology may be a feasiblesolution in a HAN. As a result, all smartdevices should be equipped with an embeddedWLAN adapter. Those devices would directlycommunicate with a WLAN home gateway thatcould also be a WLAN enabled smart meter.The authentication mechanism is performedaccording on a one-to-one basis between thesmart device and the gateway. On the otherhand, it can be claimed that 802.11 may not bea suitable communication mechanism for aHAN. This statement can be based on the sig-nificant negative consequences that wouldresult if a 802.11-based HAN network wasmaliciously attacked. For example, suppose theWLAN session is hijacked; then the attackerwould manipulate the smart devices and corre-sponding output data, and hence forward falsi-fied information to the grid. More simply,assume an attacker was able to jam a WLANcommunication by generating random data.This would cause a serious issue with the avail-ability of the HAN network, causing denial ofservice (DoS) that affects not only the func-tionality of the concerned network, but otherdependent smart grid networks as well, includ-ing NANs. Furthermore, presume that anattacker was capable of performing traffic anal-ysis on the WLAN traffic in a HAN. Conse-quently, the confidentiality of the informationwould be targeted since the attacker couldinfer HAN consumption loads of various smartdevices. In conclusion, we believe that WLAN,with its open standards, high throughput,strong home market penetration, good eco-nomics, and relatively secure communication,is a suitable choice in a HAN.

ZigBee — ZigBee is a specification for a com-munication protocol using small low-power digi-tal radios based on the IEEE 802.15.4 standard.It is more specifically known as low-rate wirelesspersonal area networks (LR-WPANs). Confiden-tiality of a Zigbee network is established thoughutilizing the AES algorithm. Moreover, frameintegrity is achieved by generating integritycodes. ZigBee devices authenticate by employingpredefined keys. Additionally, ZigBee networksprovide security countermeasures against mes-sage replays by ensuring freshness of transmittedframes. The 802.15.4 protocol is vulnerable tojamming. This threat aims to weaken the avail-ability of system services. Another threat is char-acterized by message capturing and tampering,which are difficult to avoid in LR-WPANs, sincethe cost of sufficient physical protection defeatsthe important low cost design goal of such net-works. A further threat is exhaustion; a compro-mised coordinator node can lure a large numberof nodes to associate with it by appearing to be acoordinator with high link quality. Consequently,

AMI is the key ele-

ment in smart grid

HANs [4]. It is

dubbed as the con-

vergence of the

power grid, the

communication infra-

structure and the

supporting informa-

tion architecture. It

refers to the systems

that measure, collect,

and analyze energy

usage from

advanced smart

devices



it can force all the devices to stay active for mostof the time, resulting in quick battery depletionsat those devices.

In 2007 a large stakeholder communityassembled the ZigBee Alliance to tackle theAMI and develop what is known as ZigBeeSmart Energy. Hence, this extensively advocatesthe feasibility of adopting the ZigBee technologyas a HAN communication infrastructure. As aresult, a ZigBee gateway device supporting twocommunication streams joining the utility AMIcentral database to smart devices in the HANneed to be placed and configured. The gatewaycan also act as a trust center and firewall in theZigBee network implementation to protectassets from the grid side. To complete the net-work topology, in-house smart devices equippedwith ZigBee modules should be configured andauthenticated. However, a core security threat isif, for instance, an adversary were able to com-promise a HAN coordinator ZigBee node. As aresult, this node would be able to maliciouslycontrol all aspects of other smart device nodes,tamper with their transmitted data, falsely redi-rect their communications, or even deplete theirbatteries for a complete system failure. Addi-tionally, suppose an attacker was capable of jam-ming or flooding the ZigBee HAN network.Consequently, this would trigger a drastic avail-ability problem that halts the network which willpropagate, negatively affecting all other seg-ments of the grid’s communications and func-tionality. In summary, we believe that ZigBee,with its extremely low cost (e.g., less than $10),low power consumption, unlicensed spectrumuse, and already available relatively secure“smart energy” products, is an extremely effec-tive and efficient communication choice in aHAN.

Mobile Communications and Femtocells —Femtocells are cellular network access pointsthat connect in-house user equipment (UE) tomobile operators’ core network infrastructureusing residential digital subscriber line (DSL),cable broadband connections, or optical fibers.The technologies behind femtocells are cellularsuch as Universal Mobile TelecommunicationsSystem (UMTS) and Long Term Evolution(LTE). One key driver of femtocells is thedemand for higher indoor data rates, which canbe achieved through establishing high-perfor-mance radio frequency links with a femtocell.Additionally, these devices can significantly pro-vide power savings to indoor UE since the pathloss and the required transmitting power tointerface with a femtocell are significantly lessthan to communicate with an outdoor base sta-tion. This fact renders the applicable feasibilityof mobile communications and femtocells inHANs.

In a femtocell networking environment, confi-dentiality and integrity of the transmitted dataare guaranteed by using end-to-end IPsec. More-over, authentication can be realized by usingeither the Extensible Authentication Protocol-Transport Layer Security (EAP-TLS) or theEAP-Subscriber Identity Module (SIM).

In femtocell networks, there are three mainsecurity concerns or threats. The first is charac-

terized by network and service availability. Sincethe link between the femtocell and the core net-work is IP-based, DoS and other flooding attacksare viable. The second is depicted by fraud andservice theft, where an adversary can connect toa femtocell and make illegal use of it. The thirdthreat targets privacy and confidentiality, wherethe femtocell network is subject to the samesecurity issues of regular IP-based networksincluding fabrication and modification of data.

The adoption of mobile femtocells as a HANcommunication mechanism could be a practical,reasonable, and sufficient solution. This is espe-cially true in rural HANs where other communi-cation infrastructures are unavailable but asatisfactory Internet link is accessible. Hence, ifthis architecture is realizable, smart devices(including, at least, the smart meter) should beequipped with a cellular SIM card. The authen-tication could be achieved using EAP-SIM [6]between the smart devices and the femtocell.Alternatively, smart in-house devices canauthenticate to the smart meter, and then thelatter can relay the communication to the fem-tocell. In order to enable access to the femto-cell, two access methods could be utilized:closed access and open access [7]. Issues in thedeployment of the mobile femtocell technologyin HANs could be rendered in three obstacles.First, there is concern with the use of femtocellsin homes with regard to their possible associat-ed health issues [8]. Second, there is the chal-lenge related to the ability of determiningfemtocell location. This estimation is necessaryfor smart grid operators to determine HANlocations for network planning and access con-trol, which could be hard to achieve using fem-tocells. Third, there is a security concern by gridoperators who question the transfer of sensitiveHAN data through the public Internet as atransmission medium toward the NAN andeventually the grid. In conclusion, we believethat cellular femtocells, with their relatively highprice (e.g., > $100), possible indoor healthissues, various implementation and security con-cerns, and limited device access, are not a suit-able communication choice in a HAN.

Note that the distribution part of the smartgrid (HANs and NANs) with corresponding pos-sible threats are focused on and illustrated inFig. 2.

NAN COMMUNICATION MECHANISMSThe NAN is the HAN complementary networkthat completes the distribution subpart of thesmart grid. A NAN is the next immediate tier,and its infrastructure is critical since it interre-lates and connects multiple HANs collectivelyfor the purpose of accumulating energy con-sumption information from households (theHANs) in the neighborhood and delivering thedata to the utility company. Thus, the communi-cation infrastructure responsible for such tasks isvery significant to address.

WiMAX — The IEEE 802.16 standard, referredto as Worldwide Interoperability for MicrowaveAccess (WiMAX), defines the air interface andmedium access control protocol for a wirelessmetropolitan area network (WMAN).

Femtocells are cellu-

lar network access

points that connect

in-house user equip-

ments (UEs) to

mobile operators’

core network infra-

structure using resi-

dential DSL, cable

broadband connec-

tions, or optical

fibers. The technolo-

gies behind femto-

cells are cellular such

as UMTS and LTE.

ASSI LAYOUT_Layout 1 12/21/12 12:01 PM 5


WiMAX standards define three steps to pro-vide secure communications: authentication, keyestablishment, and data encryption. This isachieved by implementing the EAP protocol, thePrivacy Key Management protocol, and theAdvanced Encryption Standard (AES) algo-rithm, respectively.

Threats in IEEE 802.16 focus on compromis-ing the radio links. Hence, the system is vulnera-ble to radio frequency (RF) jamming. WiMAXis also susceptible to scrambling attacks, wherean adversary injects RF interference while trans-mitting specific management data. This attackaffects the proper network ranging and band-width sharing capabilities. Additionally, and dueto lack of frame freshness, 802.16 is vulnerableto replay attacks.

The Smart Grid Working Group [9] acts as amajor point for utility interests in WiMAX as atechnology for smart grid networks. Thus, it pro-motes WiMAX as a core communication tech-nology for NANs. Furthermore, WiMAX is abroadband wireless last mile technology that cansupport smart grid distribution. As a result,WiMAX can be implemented between a basestation and the home gateway. The smart meterwould collect smart devices data and then for-ward them to the home gateway, which has theinteroperability property to comprehendWiMAX communication. The home gateway isin fact a subscriber station (SS) in the HAN. TheSS would collect the data from the smart meterand send it to the NAN through a WiMAX ded-icated connection. To complete the data transfertoward the utility, point-to-point, point-to-multi-point, or hybrid (multihop relay) [10] WiMAXtopologies can be implemented. The practical

feasibility of such a technology could be hin-dered by possible security misdemeanors. Forexample, since WiMAX is susceptible to trafficanalysis techniques, an adversary with maliciousintentions can retrieve HANs’ sensitive datawhile in transit through WiMAX to identifyneighborhood trends in consumption loads.Moreover, an attacker can take advantage oflack of message timeliness to launch a man-in-the-middle attack by replaying certain informa-tion from the grid or the NAN toward the HANsusing WiMAX. In summary, we believe thatWiMAX, with its high throughput, significantsmart grid standardization and working groups,backhaul media for WiFi or ZigBee in-premisesdevices, and interoperability features, is veryapplicable as a NAN communication technology.

LTE — Long Term Evolution (LTE) is a wirelesscommunication standard for a fourth generationmobile network. LTE features an all-IP flat net-work architecture, end-to-end quality of service,peak download rates nearing 300 Mb/s, andupload rates of 75 Mb/s. This makes it veryadvantageous as a NAN communication mecha-nism. LTE networks provide mutual authentica-tion between the UE and the core network byimplementing the Authentication and KeyAgreement (AKA) protocol. For radio signaling,LTE provides integrity, replay protection, andencryption between the UE and the base station(e-NB). Internet Key Exchange (IKE) coupledwith IPsec can protect the backhaul signalingbetween the e-NB and the core network [11].For user plane traffic, IKE/IPsec can similarlyprotect the backhaul from the e-NB to the corenetwork.

Figure 2. Smart grid distribution and corresponding threats.

Basestation

Home area network Neighborhood area network

Smartmeter

Homegateway

Connection-based:- RF jamming- Wireless scrambling- Eavesdropping- Message modification and injection- Protocol failures- Physical attacks and natural disasters

Device-based:- Physical attacks and natural disasters- Rogue access points- Man-in-the-middle attacks- DoS attacks- Replay attacks- Illegitimate use of services- Masquerading- Wardriving

ThreatsLTE is a wireless

communication stan-

dard for a fourth

generation mobile

network. LTE fea-

tures an all-IP flat

network architecture,

an end-to-end quali-

ty of service, peak

download rates near-

ing 300 Mbps and

upload rates of

75 Mbps. This ren-

ders it very advanta-

geous to exist as a

NAN communication

mechanism.



Threats in LTE can be divided into threemain sections. The first is characterized byattacks on the air interface. Such attacks aremainly passive, such as traffic analysis and usertracking. The second is rendered by attacks onthe e-NB. Such threats include physical tamper-ing with the e-NB, fraudulent configurationchanges, DoS attacks, and cloning of the e-NBauthentication token. The third section is char-acterized by attacks against the core network.These may include flooding and signalingattacks.

In the context of the smart grid, the adoptionof LTE as a NAN technology could be feasiblein two ways. The first is the use of the alreadyimplemented mobile network architecture ofestablished mobile network operators (MNOs)to carry out the data. This method can bereferred to as piggybacking, where smart devicesdata from HANs are piggybacked on the MNOinfrastructure as a medium to reach the NANand eventually the utility. An advantage of thisapproach is the ease of implementation andadoption since from a smart grid perspective,there is no additional needed configuration,setup, and management. The second way inwhich LTE could be adopted is by utilizing aspecialized network core architecture to transferthe data. This methodology can be realized intwo ways. The first is by implementing the notionof a mobile virtual network operator (MVNO),which means that the smart grid utility rents aportion of the traditional MNO core network forits dedicated functions. The second way is essen-tially recognized when the utility implements itsown core architecture, using the same LTE tech-nologies as the MNO, but totally decoupledfrom it.

One critical security issue that may thwartLTE as a NAN communication mechanism isthe fact that the e-NB is the main location where

users’ traffic may be compromised [11]. Hence,if various attacks on the e-NB are successful,they could give attackers full control of the e-NBand its signaling to various nodes. In this case,HANs and NANs on the grid and their commu-nications would also be compromised since insuch architecture, they play the role of sub-scribers to the e-NB in the LTE/smart grid infra-structure. To conclude, we believe that LTE,with its cost effectiveness coupled with its rela-tively rapid implementation and highly secure,available, and trustful infrastructure, is a suitableNAN communication mechanism.

A high-level illustration of the discussed com-munication mechanisms in smart grid distribu-tion networks is shown in Fig. 3.

Broadband over Power Lines — Advancedsignal processing techniques and standardiza-tion efforts performed by the European Com-mittee for Electro-technical Standardizationhave made the employment of narrowbandpower line communications (PLC) possible.The evolution of this technology gave birth tobroadband over power lines (BPL) systems.BPL offers high-speed data communicationswith minimal new infrastructure to deploy,making this technology a viable mechanism forNAN communications.

In terms of security objectives, no defaultsecurity protocols are provided by the PLCmedium access control (MAC) standards toachieve access control.

Power line channels are considered sharednetworking media; hence, external and internalattacks are feasible on such networks. Externalthreats refer to eavesdropping on exchangeddata without having access credentials. On theother hand, internal threats are performed bybenign users on the network using access cre-dentials with the intent to misuse services.

Figure 3. Distribution network-communication mechanisms.

Neighborhood area networkHome area network

Internet

Gateway/smart meter

PLC/BPL

Femtocell

Advanced signal pro-

cessing techniques

and standardization

efforts performed by

the European Com-

mittee for Electro-

technical

Standardization have

made the employ-

ment of narrow

band PLC possible.

The evolution of this

technology gave

birth to broadband

over power lines

systems.



PLC is a system that could potentially beused in NANs on the smart grid. Many stan-dards such as ITU G.Hn and IEEE P1901exist. We believe that a harmonized PLC stan-dard is possible by interoperating these sys-tems for a better implementation of the BPLfor smart grid. However, a major obstacle forsuch adoption is rendered by the fact thatelectric transformers block the transmissionfrequency of the BPL. This limits BPL to smallcoverage range within the low voltage grid(neighborhood) and requires other retransmis-sion mechanisms to allow the full data transferto the utility. From a security perspective, anattacker may be able to launch a man-in-the-middle attack by forging his/her identity, andstanding between a HAN and NAN communi-cation using BPL. Moreover, an adversary cantake advantage of the use of copper wiring inPLC to sniff the data. In summary, we consid-er that the BPL technology is unlikely toemerge as a leading broadband tool for smartgrid NANs, but instead wil l remain as anoption for NAN communication in the futuresmart grid.

In the subsequent section, we present a dis-cussion on the security framework needed toenable the above mentioned communicationtechniques to be employed for smart grid appli-cations.

SECURITY FRAMEWORK DISCUSSIONCurrently, there is a lack of adequate work insecurity schemes and frameworks for AMI, espe-cially in authentication methods. To the best ofour knowledge, there are very limited realisticapproaches [12] to solving the scalability prob-lem of smart meter authentications, regardless ofwhich communication technology is utilized.

Cryptographic methods such as digital certifi-cates require a momentous overhead in compari-son with data packet processing. In addition,cryptographic operations contribute to extensivecomputational cost. In the context of the smartgrid, a smart meter routinely sends a meterreading message within a period of 500 ms [13].Nowadays, for PKI-based schemes, generating adigital signature every 500 ms is not an issueusing a commodity computer. Conversely, for alegacy power grid that interconnects numerousbuildings, the number of meter reading messagesthat require verification by the NAN gatewaymight be noticeably larger than its capacity.Although digital signing and message verificationcan certainly achieve secure communications, webelieve that conventional cryptographic opera-tions make such security frameworks neitherscalable nor affordable.

We assert that the security frameworkrequired to enable the discussed communicationtechniques to be employed for smart grid appli-cations should be based on the following designobjectives:• Device authentication: The identity and

legality of the smart meters and their asso-ciated consumers should be verified asreceiving the proper utility services.

• Data confidentiality: The smart meter read-ings and management control messages

should be confidential to conceal both con-sumers’ and utilities’ privacy.

• Message integrity: The smart grid should beable to verify that any meter messages aredelivered unaltered in an AMI.

• Prevent potential cyber attacks: Smartmeters should be guaranteed to obtainsecure communication with the AMI net-work, even if an individual smart meter iscompromised.

• Facilitating communication overhead: Theproposed framework should be efficient interms of communication overhead and pro-cessing latency.

CONCLUSIONIn this article, we have investigated applicablecommunication mechanisms that could be adopt-ed on smart grid distribution networks. To tacklethe cyber security of such infrastructures, we havepinpointed their security objectives and threats.We have further elaborated on their practical fea-sibility in terms of their technical implementation,possible obstacles, and core security issues, andattacks on smart grid HANs and NANs.

We believe it is critical to continue discussing,designing, and implementing solutions for suchmechanisms for the purpose of enhancing thecyber security of the future electric grid andhence accomplishing consumers’ utmost trust insuch a major grid transformation.

REFERENCES[1] A. R. Metke and R. L. Ekl. “Security Technology for

Smart Grid Networks,” IEEE Trans. Smart Grid, vol. 1,no. 1, June 2010, pp. 99–107.

[2] R. Yu et al., “Cognitive Radio Based Hierarchical Com-munications Infrastructure for Smart Grid,” IEEE Net-work, vol. 25, no. 5, Sept.–Oct. 2011, pp. 6–14.

[3] J. Naruchitparames, M. H. Gunes, and C. Y. Evrenosoglu,“Secure Communications in the Smart Grid,” 2011 IEEEConsumer Commun. and Networking Conf., Jan. 2011,pp. 1171–75.

[4] U.S. Dept. of Energy, AMI System Security Require-ments, 2008, http://energy.gov/sites/prod/files/oeprod/DocumentsandMedia/14-AMI System Security Require-ments updated.pdf.

[5] A. Mishra et al., “Security Issues in IEEE 802.11 WirelessLocal Area Networks: A Survey,” Wireless Commun. andMobile Computing, vol. 4, no. 8, 2004, pp. 821–33.

[6] H. Haverinen and J. Salowey, Eds., “Extensible Authenti-cation Protocol Method for Global System for MobileCommunications (GSM) Subscriber Identity Modules(EAP-SIM)),” 2006, http://merlot.tools.ietf.org/html/rfc4186.

[7] V. Chandrasekhar, J. Andrews, and A. Gatherer, “Fem-tocell Networks: A Survey,” IEEE Commun. Mag., vol.46, no. 9, Sept. 2008, pp. 59–67.

[8] J. Zhang and G. de la Roche, Front Matter, (Wiley,2009), pp. i–xxix.

[9] The WiMAX Forum, Technical Activities and WorkingGroups), 2011, http://www.wimaxforum.org/about/technical-activities-and-working-groups.

[10] NIST, “Guide to Securing WiMAX Wireless Communi-cations: Recommendations of the National Institute ofStandards and Technology,” 2010, http://csrc.nist.gov/publications/nistpubs/800-127/sp800-127.pdf.

[11] R. Blom et al., Security in the Evolved Packet System,2011.

[12] Y. Yan, Y. Qian, and H. Sharif, “A Secure and ReliableIn-Network Collaborative Communication Scheme forAdvanced Metering Infrastructure in Smart Grid,” IEEEWCNC ’11, Mar. 2011, pp. 909–14.

[13] R. E. Castellanos and P. Millan. Design of A WirelessCommunications Network for Advanced Metering Infra-structure in A Utility in Colombia,” 2012 IEEE Colom-bian Commun. Conf. May 2012, pp. 1–6.

We believe it is criti-

cal to continue dis-

cussing, designing,

and implementing

solutions for such

mechanisms for the

purpose of enhanc-

ing the cyber security

of the future electric

grid and hence

accomplishing con-

sumers’ utmost trust

in such a major gird

transformation.



BIOGRAPHIESELIAS BOU-HARB ([email protected]) is a networksecurity researcher pursuing his Ph.D. in computer sci-ence at Concordia University, Montreal, Canada. Previ-ously, he completed his M.A.Sc. degree in informationsystems security at the Concordia Institute for Informa-tion Systems Engineering. He is also a member of theNational Cyber Forensic and Training Alliance (NCFTA),Canada. His research interests focus on the broad areaof cyber security, including operational cyber securityfor critical infrastructure, LTE 4G mobile network securi-ty, VoIP attacks and countermeasures, and cyber scan-ning campaigns.

CLAUDE FACHKHA ([email protected]) is a securityresearcher at NCFTA Canada. In 2008, he received his Bach-elor of Engineering in computer and communication fromNotre Dame University. Two years later, he received hisMaster of Engineering in information systems security fromConcordia University, where he is currently pursuing hisPh.D. degree in the Faculty of Electrical and ComputerEngineering. His current research interests are in the areasof network traffic analysis and large-scale cyber threats

MAKAN POURZANDI ([email protected]) is aresearcher at Ericsson, Canada. He received his Ph.D.degree in computer science from the University of Lyon,France, and his M.Sc. degree in computer science fromÉcole Normale Supérieure de Lyon, France. His currentresearch interests include security, cloud computing, soft-ware security engineering, cluster computing, and compo-nent-based methods for secure software development.

MOURAD DEBBABI ([email protected]) holds Ph.D.and M.Sc. degrees in computer science from Paris-XI OrsayUniversity, France. He has published more than 70 research

papers in international journals and conferences on com-puter security, formal semantics, mobile and embeddedplatforms, Java technology security and acceleration, cryp-tographic protocol specification, design, and analysis, mali-cious code detection, programming languages, typetheory, and specification and verification of safety-criticalsystems. He is a full professor and the director of the Con-cordia Institute for Information Systems Engineering, Con-cordia University, Montreal, Quebec, Canada. He has servedas a senior scientist at the Panasonic Information and Net-work Technologies Laboratory, Princeton, New Jersey; asso-ciate professor at the Computer Science Department ofLaval University, Quebec, Canada; senior scientist at Gener-al Electric Research Center, New York; research associate atthe Computer Science Department of Stanford University,Palo Alto, California; and permanent researcher at the BullCorporate Research Center, Paris, France.

CHADI ASSI ([email protected]) received his B.Eng.degree from the Lebanese University, Beirut, Lebanon, in1997 and his Ph.D. degree from the City University of NewYork (CUNY) in April 2003. He is currently an associateprofessor with the Concordia Institute for Information Sys-tems Engineering, Concordia University. Before joiningConcordia University in August 2003 as an assistant pro-fessor, he was a visiting scientist for one year with NokiaResearch Center, Boston, Massachusetts, where he workedon quality of service in passive optical access networks. Heis an Associate Editor for Wiley’s Wireless Communicationsand Mobile Computing. His research interests include opti-cal networks, multihop wireless and ad hoc networks, andsecurity. He received the prestigious Mina Rees Disserta-tion Award from CUNY in August 2002 for his research onwavelength-division multiplexing optical networks. He ison the Editorial Board of IEEE Communications Surveys &Tutorials and serves as an Associate Editor for IEEE Com-munications Letters.
